I think you should definitely apply if you are interested = in it! I just want you to be aware of the difficulty. I can be a mentor if = there aren't other people available for it.

However,= I am not sure if ruby-core is interested in this feature so it would be go= od for MRI developers to confirm whether or not such a patch would potentia= lly be accepted if it was good.

A detailed proposa= l including a plan of how you plan on investigating the existing fiber sema= ntics and creating a design plan for cross-thread resumption would be quite= helpful.

Also documenting a mid-term deliverable = would be good too. Offhand I'm not sure what that would be for this pro= ject per se, but it's important to have such a goal.

On Mon, Mar 23, 2015 at = 11:52 AM, surya pratap singh raghuvanshi <oshosurya@gmail.com> wrote:

hi tony,so should i consider that this project would not be conducted during this = gsoc ?
or should i put some more efforts in it and make a proposa= l of how i am planning to work on it ?

On= Sun, Mar 22, 2015 at 1:00 PM, Tony Arcieri <bascule@gmail.com> wrote:

Hi Surya,
I put this project on the Ideas List. It's a project = I'm potentially interested in mentoring.

Howev= er, I just want to say up front that I think it will be quite difficult. I = just sent this message to ko1 on a different ruby-core thread:

---------- Forwarded message --= --------
From:=C2=A0Tony Arcieri=C2=A0= <bascule@gmail.com>
Date: Sun, Mar 22, 2015 at 12:59 PMSubject: Re: [ruby-core:68480] Re: Looking for MRI projects for Ruby Goog= le Summer of Code 2015
To: Ruby developers <ruby-core@ruby-lang.org>

=
= On Tue, Mar 17, 2015 at 1:59 PM, SASADA Koichi=C2=A0<ko1@atdot.net&g= t;=C2=A0wrote:
Fiber captures also machine stack= . How to resume it on the other
threads? It is very interesting.

Hi Koichi,

I put t= his project on the Ideas List. It is indeed quite difficult. It's a pro= ject I'm potentially interested in mentoring, but if we chose to attemp= t it I would really love your input on it.

There a= re many potential states where resuming a fiber across threads simply will = not work, for example if the current fiber is holding a mutex. So the proje= ct would start with determining what states a fiber could be in where cross= -thread resumption is allowed.

It may simply be in= feasible without major changes to MRI. I would consider it more of a resear= ch project than anything, and would probably give it a low chance of succes= s.

--
Tony Arcieri